V.I.S.O.R. has 16 tools with consistent verb-noun naming (get_, search_, build_, list_, create_, trace_, etc.) and reasonable descriptions (avg ~120 chars). However, critical gaps undermine production readiness: (1) Most tools lack documented output schemas, LLMs cannot plan downstream calls or extract fields. (2) Parameter descriptions are minimal; many lack format/constraint guidance. (3) No error handling guidance, tools return bare failures without recovery hints. (4) Two WRITE tools (create_skill, index_workspace, switch_workspace) lack confirmation/dry-run patterns. (5) Tool definitions appear inferred from FastMCP decorators without explicit schema visibility in source. Naming is strong; descriptions are adequate but not LLM-optimized; schemas are largely absent.
Build ranked context for a natural language query using skill-based scoring strategies
Create a custom skill strategy for context building
Identify potentially dead code (unused functions, classes, variables) in the codebase
Generate a high-level architectural map showing major components and their relationships
Trace the dependency chain from a source file or symbol, showing all direct and transitive dependencies
Get the full content and metadata of a specific file
Get detailed context for a specific symbol (function, class, variable) including definition, usage, and documentation
Output schemas not documented. LLMs cannot infer what fields tools return, forcing guesswork on downstream tool chaining and field extraction. All 16 tools lack visible return type documentation.
Parameter descriptions lack format/constraint guidance. E.g., 'depth' in get_dependency_chain and get_architecture_map has no min/max bounds; 'search_type' enum is present but other params lack similar constraints. LLMs cannot validate inputs without explicit ranges.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 53 | <=2025-11-25 | v2 |
Get telemetry and usage statistics for the V.I.S.O.R. server
Basic health-check endpoint confirming the server is alive.
Analyze the impact of changes to a file or symbol by tracing all downstream dependents
Trigger indexing of the workspace to build the code graph and embeddings
List all available skill strategies for context building
List all registered workspaces and their statistics
Search the codebase for symbols, patterns, or text using semantic and exact matching
Switch the active workspace context
Trace the shortest architectural path between two files or symbols in the dependency graph
WRITE tools (create_skill, index_workspace, switch_workspace) lack confirmation/dry-run patterns. Agents can accidentally create duplicate skills or reindex large workspaces without safeguards. No error recovery guidance provided.
No error handling guidance. Tools return failures without actionable recovery hints. E.g., if search_codebase finds no results or get_file_context fails on invalid path, LLM receives no guidance on next steps (retry, try alternative tool, ask user).
Tool definitions inferred from FastMCP decorators without explicit schema visibility in source code. Cannot verify full schema compliance.